I have a 40 year monthly rainfall time series with 10 years gap in that.
I have two questions:
1- how many years we can reconstruct the missing data? is there any role for this?
2- what is the best reconstruction method for filling the gaps in a meteorological time series.
Sincerely
Dear Majid,
see also this discussion: https://www.researchgate.net/post/Where_can_I_freely_download_global_climate_data_sets.
Best regards
Nils
https://www.researchgate.net/post/monthly_rainfall_data_missing
https://www.researchgate.net/post/How_to_fill_missing_measured_temperature_data
https://www.researchgate.net/post/Up_to_what_point_i_can_fill_missing_precipitation_data
https://www.researchgate.net/post/How_might_I_fill_missing_data_of_river_flow_and_temperature_data_hydrological_data
https://www.researchgate.net/publication/285501813_Analyse_par_les_vecteurs_regionaux_de_la_variabilite_spatio-temporelle_des_precipitations_annuelles_liees_a_la_mousson_africaine?_iepl%5BgeneralViewId%5D=6KW43A1wOAGE1KJJl56TnsuPruKtwo05uEjH&_iepl%5Bcontexts%5D%5B0%5D=searchReact&_iepl%5BviewId%5D=JbhyzRSEyH1BNiKGoZ9J3FlTliwOUfNhZ3NQ&_iepl%5BsearchType%5D=publication&_iepl%5Bdata%5D%5BcountLessEqual20%5D=1&_iepl%5Bdata%5D%5BinteractedWithPosition9%5D=1&_iepl%5Bdata%5D%5BwithEnrichment%5D=1&_iepl%5Bposition%5D=9&_iepl%5BrgKey%5D=PB%3A285501813&_iepl%5BtargetEntityId%5D=PB%3A285501813&_iepl%5BinteractionType%5D=publicationTitle
Because of coarse resolution and atmospheric obscurity to retrieve the rainfall data from satellites, the available rainfall data already has a significant issues. You may fill in the gaps in one way or another, but I remind you that you are filling just trash and no confidence. However, you may fill in very small gaps with coupled meteorological models but still with a significant uncertainty. Hope this helps!
Dear Majid Dehghani,
As far as I know, there are many techniques which can be used to fill gaps in time series. Those techniques range from statistical-based regression approaches to ANN-based techniques.
I can direct you to a recent paper where the authors performed the reconstruction of a large streamflow dataset over west and central africa using two different approaches: Multiple Imputation by Chained Equations (MICE), that is to say, a chain of regression equations, which were later then compared to a Random-Forest (ANN-based) approach.
Find the paper below.
Article Trend and variability in a new, reconstructed streamflow dat...
convert rainfall data to spatial and do the Interpolation technique in ArcGIS (think about the Time Step format)
https://scholar.google.fr/scholar?hl=fr&as_sdt=0%2C5&q=filling+gaps+rainfall&btnG=
I strongly suggest that you avoid any fully automatic reconstruction method. As as I said it really depends on the time step, and of the time of the year.
If you have several stations around or not.
You can also build a theoretical regional daily rainfall regime from existing neighbouring stations, and use it as a reference to apply anomalies
If you have available data from neighboring stations, I recommend to use them for gapfiiling. For the method, I would recommend group method of data handling or even simple method of regression analysis.
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Atmosphere